Keeping A Good Laboratory Record Book

Dr Roland Smith.
As part of first year laboratory course you are required to keep a lab book. The
keeping of detailed records of your work is essential both as a student and as a
professional scientist. More generally, recording what you did, what you found, who
you spoke to, how much money you spent etc are critical aspects of good business
practice outside the field of scientific research. The aim of this document is to
highlight why lab books are important and how you should go about keeping them.

1. Why keep a lab book?

There are a number of extremely good reasons for keeping a suitably detailed and
robust lab book. Some of the most important of these are:
Experiments may take months or years to complete, and analysing data and writing up
your results is impossible without decent records. You will never remember all the
critical details without a written record. Indeed, you often dont find out what the
critical points are until well after an experiment has been finished.
On occasion very significant amounts of prestige or money can be involved in who
did what and when. For this reason lab books are legal documents of record and in
industry it is very common for lab books to be counter signed by your supervisor
every week to set up a well defined paper trail.
Both industrial and academic research projects are often carried out in large teams.
People leave projects, go on holiday, have accidents, equipment fails etc. For this
reason a permanent record of work done, experimental details, operating procedures
etc need to be available, and, most critically, be understandable to other people.
Lastly, on a more personal level, your demonstrators will mark your lab book and
your degree results will to some extent depend on how well your lab book is kept.

2. What should a lab book be?

A lab book should be a real time record of what you do in the lab as it happens. It has
to be:
Complete

The lab book should contain everything, including the mistakes. If

something is wrong, cross it out in a way that still lets you read it.

Tough

No loose papers, graphs etc. Securely glue or staple things in. In a

real lab environment expect to have your lab booked dropped, kicked,
used to prop up equipment, stained with coffee and so on.

Clear

The lab book doesnt have to be pretty, but it must be legible and
contain sufficient detail (headings, figure captions, table titles, units
etc) to make the contents understandable years later.

True

The lab book is a record of what you see and do. Dont embellish it,
dont draw plots of raw data the way you think they should look, draw
them the way they are. Dont throw away bad points and things
which dont appear to agree with theory at the time.
1

3. What should your lab book include?

A lab book does not have to be carefully structured and presented to the same degree
as a formal lab report. However there are a number of key thinks that need to be
included. A complete record of an experiment or project will probably contain the
following points:
Heading

A suitably descriptive title for the work.

Date and time of start

(and additional dates for each day you work on it).

Aims

What is it that you want to do?

Background

References to lab scripts, text books, manuals etc

Description of set up

Best accomplished with the aid of simple figures. What

apparatus did you use, how was it put together, what
was it supposed to do?

Initial set up and

preliminary data

It is very common to set out to do a quick, sloppy job

on an experiment first time round, with the expectation
that this will show you what you need to go back and do
well to get your real data.

Review

Almost nothing worthwhile works properly the first

time. This is one of the key lessons that lab work will
teach you. What went wrong first time round? How are
you going to fix it?

Definitive data

Once you understand the experiment you can home in

on the key results you need and do these with particular
care. Usually a clear, well labelled table with units,
errors on individual readings etc.

Data analysis including

error calculations

Any result from an experiment is meaningless without

consideration of the errors involved.

Comparisons and
conclusions

Did the experiment agree with theory, work as

expected, need doing again etc.

Summary

What was the final result, what worked well, what

didnt, what could be improved in the future? This part
is particularly critical for the short lab experiments

Closure

Be clear where one thing finishes and another starts.

Note: Some people reserve the last page of the lab book for an index of the books
contents which grows as the book is used. To do this you need to number the pages in
your lab book. Even if you dont use an index, numbered pages can be very useful for
cross-referencing other material in the book. e.g. you might re-do an experiment on
Nov 17th 2002 that you first attempted on Oct 24th 2002. On Nov 17th you might write
on page 60: I have found a systematic error which must have affected the results
taken on Oct 24th 2002 (see page 35). Cross reference this by adding a note to your
original work on page 35 saying: Nov 17, 2002, See note on page 60.

4. Some notes on computing experiments.

The guidelines above are primarily aimed at practical experiments such as the
electronics or optics sections of first year lab. The computing lab involves somewhat
different skills and is typically documented rather differently in your lab book. You
should still keep a lab book for the computing experiment and it should include a
record of the problems you are trying to solve, general methods employed by routines
etc along with hard copies of program listings and example print outs. However
much of the information that would be explicitly included in a lab book in a
traditional experiment will appear as typed comments in the code rather than notes
and figures written by hand.

5. Neatness versus speed: Getting the balance correct.

A lab book is a real time record of what you do in the lab. However there is a danger
that if you spend too much time on the lab book and not enough on the experiment
you will not have much useful to say in it. For this reason you need to get the balance
right between time spent keeping the lab book and time spent actually getting the
experiment done. A quick, well-annotated freehand sketch done in two minutes can
convey just as much information as a careful engineering drawing done to scale with a
ruler, set square and compass. On the other hand a column of numbers alone will be
useless a couple of days after the experiment unless accompanied by a brief
description of what is being recorded and why. Lets look at a few examples:
The figure of an oscilloscope trace below carries some essential information, but is
over complicated. There is no point in including an explicit scale for example when
a simple note of peak signal height and peak-to-peak separation is just as
informative. Never use a ruler to draw the actual scope trace, you tend to miss
important details.

The simple sketch overleaf is quicker to draw, conveys more information and is
actually easier to read important points off. Freehand drawing of the scope trace
makes it much easier to include things like the ringing effect seen below. The text
on the figure identifies it with a specific measurement and also highlights some queries
that might need looking at again.
3

This list of numbers is perfectly legible but I

have no idea what it means.

This set is a bit messier, but accompanied by a

simple heading and description which will
actually make it useful when it comes to writing
up the experiment. Errors have been corrected
in such a way as to leave the wrong result
clear. This is a critical point, experimental
physics almost never happens in a perfectly
logical order in the real world. Sometimes you
make real mistakes, on some occasions
(particularly late at night) what you think is
wrong at the time is actually correct and will
be needed later. For this reason never remove
pages from your lab book, tipex things out etc.

Original document written October 2000. Last updated

September 2005.